Problem behaviors have included a number of different constructs, such as delinquency, promiscuity, aggression, that have all been shown to relate to the onset and progression of substance use. However, little is known about the substance-specific relationship between problem behaviors and substance use where initial levels of problem behaviors predict change over time of three forms of substances (i.e. alcohol, tobacco, and marijuana). The overall goal for the proposed study is to fill in these gaps in the etiology of substance use and problem behaviors in this salient developmental period of adolescence, focusing on alternative reinforcers (i.e. dancing, dating, sports, being involved in clubs) as a possible mechanism that may account for changes in behaviors over time. The theoretical model proposed hypothesizes that adolescents with higher levels of problem behaviors have learned that they can experience substantial changes in baseline levels of arousal only in response to highly stimulating behaviors, such as stealing, skipping school, getting into fights. These adolescents with higher levels of problem behaviors may experience environments with healthy alternative non-drug reinforcers as less stimulating. The hopeful impact of this project is that it will better inform clinicians and policy interventions to supportthe increasing rates of substance use in the United States. As more adolescents are engaging in a variety of substances, additional insights will be gained in the specific temporal relationship between problem behaviors and substance use. Furthermore, if alternative reinforcers are shown to mediate the relationship between problem behaviors and substance use, this would have an impact on behavioral activation interventions that can teach adolescents other pleasurable activities that can mitigate their substance use. The project will utilize a rich data et from the Happiness and Health Study. This study comprises a sample of 3,383 adolescents across 10 different high schools in the southern California area. The first wave of data collection began in Fall 2013 when the students were beginning 9th grade and will continue with semiannual assessments until Spring 2017 prior to the student's graduation. Furthermore, using innovative statistical techniques, such as structural equation models with latent difference scores, will provide a more nuanced understanding of behavior changes across time. This sample also has a large percentage of Hispanics (47.4%) and will allow us to examine possible ethnic differences in consumption patterns as well as possible mechanisms towards use. By integrating the rich panel of longitudinal psychosocial, substance use, and problem behavior data from the Happiness and Health study to examine changes in use over time, the proposed project offers an exceptional opportunity for training and to gain depth of insight into factors related to substance use onset and progression and possible points of intervention.